Modernising core banking with Generative AI
When advantages turn into results
Scalable core banking modernisation with lower risk
The bank proved it can modernise large portions of its core systems using a scalable, repeatable model that reduces long-term program risk.
GenAI accelerates large-scale legacy modernisation
GenAI accelerated modernisation tasks manual methods cannot support at such a scale, including documentation, testing and code refinement across millions of lines.
Faster time to market with cloud-ready core systems
The transformation strengthens the global financial institution’s ability to compete with digital native challengers by improving time to market, reducing infrastructure costs and enabling cloud-ready scale.
GenAI unlocks large-scale legacy transformation by automating what manual methods can't handle—comprehensive documentation, testing, and code refinement at enterprise scale.
Challenge
Why modern banks need a new approach to IT modernisation
Many of the bank’s foundational systems were designed decades ago for branch and back-office operations. Today, these systems must always support on digital services, mobile channels and new patterns of customer interaction. As volumes grew, the mainframe became increasingly expensive to scale and maintain.
The bank also faced a structural challenge shared across the industry: legacy workloads were tightly interconnected, making change difficult and time-consuming. With COBOL talent becoming scarce, maintaining and extending these systems created operational and delivery risk. Manual approaches to legacy modernisation were not viable for millions of interdependent lines of code. The global financial institution required a modern architecture built for cloud scale, flexibility and long-term resiliency.
Engagement
AI-enhanced modernisation with GFT and GenAI: Combining deterministic conversion and AI in software development
GFT and the international bank adopted a structured modernisation approach that blended deterministic code conversion with AI-driven acceleration through GenAI – including among others Wynxx, GFT’s AI-driven platform designed to accelerate the Software Development Life Cycle (SDLC). The work came together across four focus areas that shaped how modernisation unfolded in practice.
- Modernisation Strategy: The work began with detailed discovery and analysis to understand dependencies, complexity and architectural requirements. Deterministic tools converted COBOL to Java, preserving core business logic while creating the basis for distributed operation.
- Role of Generative AI: GenAI provided essential automation across tasks that would otherwise require extensive manual analysis. It generated documentation for legacy logic, created test scenarios to support end-to-end validation, and improved the readability and maintainability of converted Java. It also contributed to batch modernisation by supporting Python-based runtime patterns. Across each of these areas, teams moved faster and maintained consistency at scale.
- Architectural Evolution: GFT evolved the distributed architecture to support this cloud-native transformation, enhancing execution contexts, data access, queue handling, and runtime components.
- A Unified Pattern: This work delivered a modernisation approach that this international bank can apply across multiple initiatives, ensuring consistency as the program expands.
Benefit
A proven path to bank-wide modernisation
The project confirmed that modernisation at this scale is both achievable and sustainable with the right combination of automation, architecture evolution, and AI support. The client validated that a significant portion of its legacy estate can be migrated using this model, reducing long-term program risk and creating alignment across future modernisation efforts.
AI-assisted documentation, testing and refinement improved modernisation velocity and clarity for engineering teams. The updated architecture provides greater scalability, resiliency and efficiency, supporting faster time to market and lower infrastructure costs. The work has been recognised internally at both the global financial institution and GFT as a flagship modernisation effort.
With this foundation in place, the bank is positioned to expand modernisation across additional systems and explore further opportunities for AI-assisted refinement of modernised code.
Conclusion
Results from an AI-accelerated IT modernisation program
Global modernisation programs require more than code migration. They depend on automation that reduces manual effort, architectural evolution that supports scale, and AI-driven support that accelerates complex tasks without compromising accuracy or governance. The project demonstrates how AI in software development can enable modernisation at a level that legacy methods cannot achieve.
By partnering with GFT, the banking client has established a future-ready platform and a repeatable banking modernisation path, strengthening the bank’s ability to innovate, scale, and deliver reliable digital services to millions of customers worldwide.
